Evaluation of mediastinal lymph node segmentation of heterogeneous CT data with full and weak supervision.

Accurate lymph node size estimation is critical for staging cancer patients, initial therapeutic management, and assessing response to therapy. Current standard practice for quantifying lymph node size is based on a variety of criteria that use uni-directional or bi-directional measurements. Segmentation in 3D can provide more accurate evaluations of the lymph node size. Fully convolutional neural networks (FCNs) have achieved state-of-the-art results in segmentation for numerous medical imaging applications, including lymph node segmentation. Adoption of deep learning segmentation models in clinical trials often faces numerous challenges. These include lack of pixel-level ground truth annotations for training, generalizability of the models on unseen test domains due to the heterogeneity of test cases and variation of imaging parameters. In this paper, we studied and evaluated the performance of lymph node segmentation models on a dataset that was completely independent of the one used to create the models. We analyzed the generalizability of the models in the face of a heterogeneous dataset and assessed the potential effects of different disease conditions and imaging parameters. Furthermore, we systematically compared fully-supervised and weakly-supervised methods in this context. We evaluated the proposed methods using an independent dataset comprising 806 mediastinal lymph nodes from 540 unique patients. The results show that performance achieved on the independent test set is comparable to that on the training set. Furthermore, neither the underlying disease nor the heterogeneous imaging parameters impacted the performance of the models. Finally, the results indicate that our weakly-supervised method attains 90%- 91% of the performance achieved by the fully supervised training.

Drinking history associations with regional white matter volumes in alcoholic men and women.

BACKGROUND: Alcoholism has been repeatedly associated with gray and white matter pathology. Although neuroimaging has shown alcoholism-related brain volume reductions and axonal compromise, the integrity of white matter volumes in chronic alcoholism has been challenging to measure on a regional level. METHODS: We first examined the effects of alcoholism on cerebral white matter volumes by lobar and gyral subdivisions in 42 abstinent alcoholics and 42 control participants (split evenly by gender). We also examined cerebellar white matter and regions of the corpus callosum, as well as ventricular volumes. Next, relationships between white matter and ventricular volumes with measures of drinking patterns were assessed. Finally, an examination of early versus late abstinence was conducted. Within each examination, gender effects were explored. RESULTS: Differences in regional white matter volumes between alcoholics and controls were observed primarily in the corpus callosum, with a stronger group difference among men than women. Years of heavy drinking had a strong negative impact on frontal and temporal white matter among alcoholic women, and on the corpus callosum among alcoholic men. Quantity of alcohol consumption was associated with smaller corpus callosum and larger ventricular volumes among alcoholic women, whereas abstinence duration was associated with larger corpus callosum volume among alcoholic men. Preliminary data indicated that alcoholic women showed stronger positive associations between sobriety duration and white matter volume than men within the first year of abstinence, whereas men showed this association more so than women after 1 year of abstinence. CONCLUSIONS: Effects of drinking history on white matter and ventricular volumes vary by gender, with alcoholic women showing greatest sensitivity in frontal, temporal, ventricular, and corpus callosum regions, and alcoholic men showing effects mainly in the corpus callosum. Preliminary results indicate that recovery of white matter volume may occur sooner for women than for men.

Open Health Imaging Foundation Viewer: An Extensible Open-Source Framework for Building Web-Based Imaging Applications to Support Cancer Research.

PURPOSE: Zero-footprint Web architecture enables imaging applications to be deployed on premise or in the cloud without requiring installation of custom software on the user's computer. Benefits include decreased costs and information technology support requirements, as well as improved accessibility across sites. The Open Health Imaging Foundation (OHIF) Viewer is an extensible platform developed to leverage these benefits and address the demand for open-source Web-based imaging applications. The platform can be modified to support site-specific workflows and accommodate evolving research requirements. MATERIALS AND METHODS: The OHIF Viewer provides basic image review functionality (eg, image manipulation and measurement) as well as advanced visualization (eg, multiplanar reformatting). It is written as a client-only, single-page Web application that can easily be embedded into third-party applications or hosted as a standalone Web site. The platform provides extension points for software developers to include custom tools and adapt the system for their workflows. It is standards compliant and relies on DICOMweb for data exchange and OpenID Connect for authentication, but it can be configured to use any data source or authentication flow. Additionally, the user interface components are provided in a standalone component library so that developers can create custom extensions. RESULTS: The OHIF Viewer and its underlying components have been widely adopted and integrated into multiple clinical research platforms (e,g Precision Imaging Metrics, XNAT, LabCAS, ISB-CGC) and commercial applications (eg, Osirix). It has also been used to build custom imaging applications (eg, ProstateCancer.ai, Crowds Cure Cancer [presented as a case study]). CONCLUSION: The OHIF Viewer provides a flexible framework for building applications to support imaging research. Its adoption could reduce redundancies in software development for National Cancer Institute-funded projects, including Informatics Technology for Cancer Research and the Quantitative Imaging Network.

Alcoholism and dampened temporal limbic activation to emotional faces.

BACKGROUND: Excessive chronic drinking is accompanied by a broad spectrum of emotional changes ranging from apathy and emotional flatness to deficits in comprehending emotional information, but their neural bases are poorly understood. METHODS: Emotional abnormalities associated with alcoholism were examined with functional magnetic resonance imaging in abstinent long-term alcoholic men in comparison to healthy demographically matched controls. Participants were presented with emotionally valenced words and photographs of faces during deep (semantic) and shallow (perceptual) encoding tasks followed by recognition. RESULTS: Overall, faces evoked stronger activation than words, with the expected material-specific laterality (left hemisphere for words, and right for faces) and depth of processing effects. However, whereas control participants showed stronger activation in the amygdala and hippocampus when viewing faces with emotional (relative to neutral) expressions, the alcoholics responded in an undifferentiated manner to all facial expressions. In the alcoholic participants, amygdala activity was inversely correlated with an increase in lateral prefrontal activity as a function of their behavioral deficits. Prefrontal modulation of emotional function as a compensation for the blunted amygdala activity during a socially relevant face appraisal task is in agreement with a distributed network engagement during emotional face processing. CONCLUSIONS: Deficient activation of amygdala and hippocampus may underlie impaired processing of emotional faces associated with long-term alcoholism and may be a part of the wide array of behavioral problems including disinhibition, concurring with previously documented interpersonal difficulties in this population. Furthermore, the results suggest that alcoholics may rely on prefrontal rather than temporal limbic areas in order to compensate for reduced limbic responsivity and to maintain behavioral adequacy when faced with emotionally or socially challenging situations.

Alcoholism gender differences in brain responsivity to emotional stimuli.

Men and women may use alcohol to regulate emotions differently, with corresponding differences in neural responses. We explored how the viewing of different types of emotionally salient stimuli impacted brain activity observed through functional magnetic resonance imaging (fMRI) from 42 long-term abstinent alcoholic (25 women) and 46 nonalcoholic (24 women) participants. Analyses revealed blunted brain responsivity in alcoholic compared to nonalcoholic groups, as well as gender differences in those activation patterns. Brain activation in alcoholic men (ALCM) was significantly lower than in nonalcoholic men (NCM) in regions including rostral middle and superior frontal cortex, precentral gyrus, and inferior parietal cortex, whereas activation was higher in alcoholic women (ALCW) than in nonalcoholic women (NCW) in superior frontal and supramarginal cortical regions. The reduced brain reactivity of ALCM, and increases for ALCW, highlighted divergent brain regions and gender effects, suggesting possible differences in the underlying basis for development of alcohol use disorders.

Immune-related tumour response assessment criteria: a comprehensive review.

Growing emphasis on precision medicine in oncology has led to increasing use of targeted therapies that encompass a spectrum of drug classes including angiogenesis inhibitors, immune modulators, signal transduction inhibitors, DNA damage modulators, hormonal agents etc. Immune therapeutic drugs constitute a unique group among the novel therapeutic agents that are transforming cancer treatment, and their use is rising. The imaging manifestations in patients on immune therapies appear to be distinct from those typically seen with conventional cytotoxic therapies. Patients on immune therapies may demonstrate a delayed response, transient tumour enlargement followed by shrinkage, stable size, or initial appearance of new lesions followed by stability or response. These newer patterns of response to treatment have rendered conventional criteria such as World Health Organization and response evaluation criteria in solid tumours suboptimal in monitoring changes in tumour burden. As a consequence, newer imaging response criteria such as immune-related response evaluation criteria in solid tumours and immune-related response criteria are being implemented in many trials to effectively monitor patients on immune therapies. In this review, we discuss the traditional and new imaging response criteria for evaluation of solid tumours, review the outcomes of various articles which compared traditional criteria with the new immune-related criteria and discuss pseudo-progression and immune-related adverse events.

LesionTracker: Extensible Open-Source Zero-Footprint Web Viewer for Cancer Imaging Research and Clinical Trials.

Oncology clinical trials have become increasingly dependent upon image-based surrogate endpoints for determining patient eligibility and treatment efficacy. As therapeutics have evolved and multiplied in number, the tumor metrics criteria used to characterize therapeutic response have become progressively more varied and complex. The growing intricacies of image-based response evaluation, together with rising expectations for rapid and consistent results reporting, make it difficult for site radiologists to adequately address local and multicenter imaging demands. These challenges demonstrate the need for advanced cancer imaging informatics tools that can help ensure protocol-compliant image evaluation while simultaneously promoting reviewer efficiency. LesionTracker is a quantitative imaging package optimized for oncology clinical trial workflows. The goal of the project is to create an open source zero-footprint viewer for image analysis that is designed to be extensible as well as capable of being integrated into third-party systems for advanced imaging tools and clinical trials informatics platforms. Cancer Res; 77(21); e119-22. ©2017 AACR.

Comparison of performance of various tumour response criteria in assessment of regorafenib activity in advanced gastrointestinal stromal tumours after failure of imatinib and sunitinib.

PURPOSE: To compare performance of various tumour response criteria (TRCs) in assessment of regorafenib activity in patients with advanced gastrointestinal stromal tumour (GIST) with prior failure of imatinib and sunitinib. METHODS: Twenty participants in a phase II trial received oral regorafenib (median duration 47 weeks; interquartile range (IQR) 24-88) with computed tomography (CT) imaging at baseline and every two months thereafter. Tumour response was prospectively determined on using Response Evaluation Criteria in Solid Tumours (RECIST) 1.1, and retrospectively reassessed for comparison per RECIST 1.0, World Health Organization (WHO) and Choi criteria, using the same target lesions. Clinical benefit rate [CBR; complete or partial response (CR or PR) or stable disease (SD)≥16 weeks] and progression-free survival (PFS) were compared between various TRCs using kappa statistics. Performance of TRCs in predicting overall survival (OS) was compared by comparing OS in groups with progression-free intervals less than or greater than 20 weeks by each TRC using c-statistics. RESULTS: PR was more frequent by Choi (90%) than RECIST 1.1, RECIST 1.0 and WHO (20% each), however, CBR was similar between various TRCs (overall CBR 85-90%, 95-100% agreement between all TRC pairs). PFS per RECIST 1.0 was similar to RECIST 1.1 (median 44 weeks versus 58 weeks), and shorter for WHO (median 34 weeks) and Choi (median 24 weeks). With RECIST 1.1, RECIST 1.0 and WHO, there was moderate concordance between PFS and OS (c-statistics 0.596-0.679). Choi criteria had less favourable concordance (c-statistic 0.506). CONCLUSIONS: RECIST 1.1 and WHO performed somewhat better than Choi criteria as TRC for response evaluation in patients with advanced GIST after prior failure on imatinib and sunitinib.

Anatomic and metabolic evaluation of peripheral nerve sheath tumors in patients with neurofibromatosis 1 using whole-body MRI and (18)F-FDG PET fusion.

PURPOSE: Malignant peripheral nerve sheath tumors (MPNSTs) are the leading cause of death for patients with neurofibromatosis type 1 (NF1). Identification of hypermetabolic lesions on PET may help identify patients at risk for MPNST. The objective of this study was to identify clinical and MRI-derived variables that predicted increased metabolic activity of neurofibromas in NF1 patients as determined by PET. METHODS: This prospective study included NF1 patients with neurofibromas of 5 cm in diameter or greater. All patients underwent whole-body MRI and F-FDG PET imaging. Tumor volume was calculated from the MR scans using a semiautomated 3-dimensional segmentation method. SUVmax's were calculated to quantify metabolic activity. Logistic regression analyses were performed to determine the relationship among SUVmax, tumor volume, location (extremity vs trunk), type (plexiform vs circumscribed), depth (superficial vs deep), patient age, and whole-body tumor burden. RESULTS: A total of 311 neurofibromas were identified in 19 NF1 patients (mean age, 38 years; range, 19-58 years). One extreme outlier was excluded from analysis. Whole-body tumor volumes ranged from 0.4 to 1182.4 mL. Fifty of 310 tumors were FDG-avid on PET (16%) with median SUVmax of 2.2 (range, 0.4-9.6). Metabolic activity (SUVmax >2.5) correlated with tumor location (deep > superficial, trunk > extremity) in tumors with PET avidity. CONCLUSIONS: In NF1 patients with neurofibromas of 5 cm or greater, the majority of internal tumors are not metabolically active on PET. Tumors with increased metabolic activity as defined by SUVmax greater than 2.5 (ie, suggestive of MPNST) are more likely to be deep and located within the trunk.

Three-dimensional volumetrics for tracking vestibular schwannoma growth in neurofibromatosis type II.

OBJECTIVE: Vestibular schwannomas (VS) are common, benign, VIIIth cranial nerve tumors. Treatment in patients with the genetic disorder neurofibromatosis type II (NF2) is complicated by their development of bilateral VS and risk of complete deafness. Intervention decisions consider several clinical factors including tumor size and growth rate evaluated using magnetic resonance imaging. The current study evaluated the relative sensitivity of volumetric versus linear diameter measurement for assessing VS growth rate and progression. METHODS: Retrospective analysis was performed on 43 magnetic resonance imaging scans acquired longitudinally (range, 2-7 yr) from 10 patients with NF2. Fifteen VS were measured (five patients had unilateral lesions meeting inclusion criteria) using both maximum linear diameter and semiautomated volumetric analysis. Progression was defined according to Response Evaluation Criteria in Solid Tumors and its volumetric (cubed linear) equivalent. Measurement techniques were compared by assessing sensitivity to lesion growth. RESULTS: Volumetric measures were significantly more sensitive to VS growth, both for total change and change per year percentages; cubed linear growth measures (proportional to volume growth) underestimated volume growth by 50%. Seven lesions showed progression on volumetric analysis, but two of these did not show progression based on linear measures. Thus, for 29% of lesions showing progression based on volume, linear measures did not detect progression. CONCLUSION: Linear measurements underestimate VS growth rate compared with volumetric measures in NF2 patients. These results provide clear, quantitative proof that diameter measures are not as sensitive to change as volumetric measurements and that volumetric measurements should be strongly considered when making VS treatment decisions.

Brain activation during semantic processing in autism spectrum disorders via functional magnetic resonance imaging.

Language and communication deficits are core features of autism spectrum disorders (ASD), even in high-functioning adults with ASD. This study investigated brain activation patterns using functional magnetic resonance imaging in right-handed adult males with ASD and a control group, matched on age, handedness, and verbal IQ. Semantic processing in the controls produced robust activation in Broca's area (left inferior frontal gyrus) and in superior medial frontal gyrus and right cerebellum. The ASD group had substantially reduced Broca's activation, but increased left temporal (Wernicke's) activation. Furthermore, the ASD group showed diminished activation differences between concrete and abstract words, consistent with behavioral studies. The current study suggests Broca's area is a region of abnormal neurodevelopment in ASD, which may be linked with semantic and related language deficits frequently observed in ASD.

Maturation of cognitive processes from late childhood to adulthood.

To characterize cognitive maturation through adolescence, processing speed, voluntary response suppression, and spatial working memory were measured in 8- to 30-year-old (N = 245) healthy participants using oculomotor tasks. Development progressed with a steep initial improvement in performance followed by stabilization in adolescence. Adult-level mature performance began at approximately 15, 14, and 19 years of age for processing speed, response inhibition, and working memory, respectively. Although processes developed independently, processing speed influenced the development of working memory whereas the development of response suppression and working memory were interdependent. These results indicate that processing speed, voluntary response suppression, and working memory mature through late childhood and into adolescence. How brain maturation specific to adolescence may support cognitive maturation is discussed.